Speech and language deficits are often the very first symptoms of neurodegenerative diseases, such as frontotemporal lobar degeneration, corticobasal degeneration and Alzheimer[unreadable]s disease. When speech and language deficits remain isolated for at least two years, the term [unreadable]primary progressive aphasia[unreadable] (PPA) applies. Characterization of the linguistic impairments in this initial stage of the disease can provide critical information for early differential diagnosis. However, speech and language functions are rarely assessed in the setting of dementia clinics and the opportunity is often lost. Moreover, data from large-scale longitudinal clinical and neuroimaging studies of PPA are not available to guide clinicians in making diagnostic, prognostic and therapeutic decisions. We propose a detailed, four-year longitudinal study of the linguistic, anatomical and pathological features of over 200 patients with primary progressive aphasia. The accrual of this large cohort will be possible because the UCSF Memory and Aging Center is the site of an NIH-funded research program in atypical and early-onset dementias. This program will also provide general neuropsychological, neurological, neuropathological and most neuroimaging data. In aim 1 we will correlate scores in detailed language tests and regional gray matter volumes; in aim 2 diffusion tensor imaging and functional MRI will be used to investigate alteration in structural and functional connectivity in the language network in PPA; and in aim 3 we will associate post-mortem histopathological and in-vivo cortical amyloid findings with longitudinal clinical, cognitive and anatomical data to find patterns of progression typical of each neurodegenerative disease. Evidence gathered from this research will increase our knowledge about the neural basis of speech and language functions and provide crucial data for the diagnosis of neurodegenerative diseases in their early stages, when treatment can be most effective